In recent years, with the intelligence of intelligent terminals and the rapid development of network technologies, there are more and more information processing technologies of downloading information from a server and applying the information to various Applications (APP), and these information processing technologies can satisfy the needs in user's daily life and work.
An information processing application scenario can be a scenario for a common Karaoke APP on the intelligent terminal; such an APP installed on the intelligent terminal can play for the user an audio file, like a piece of accompaniment music corresponding to a song, which is downloaded from the server, acquire voice information of user and recognize it, and upload a recognition result or a comparison result which is obtained by further comparing the recognition result with preset standard reference information to the server for others to listen and comment.
Such an APP can only play all the downloaded audio files while acquiring the voice information of user, for example, the user sings and records a whole song, so there are following two problems:
1. using the traditional art cannot truncate a part of content of the audio file to play, so the larger the amount of acquired information is, the more complex the corresponding recognition effect is, and the longer the time spent on recognition is;
2. based on his/her own preference and voice, the user sings with a high pitch or a low pitch, which causes that the recognition result does not match the preset standard reference information, so the best vocal performance result cannot be obtained; that is, the user may be interested only in a part of content of the audio file, then it is only needed to truncate a part of content of the audio file, but using the traditional art cannot truncate a part of content of the audio file to play.